The purpose of this project is to examine the relationship of alcohol use to condom use from a methodological perspective. Studies on the link between drinking and risky sex have yielded inconsistent results that vary depending upon the method used to analyze the relationship. Although correlational studies show that heavier drinkers have more unprotected sex, studies of specific sexual encounters find that drinking reduces condom use only among adolescents, and studies that incorporate multiple sexual encounters per person show no relationship of drinking to condom use. Thus, measuring alcohol use and sexual behavior at different levels-the person level, the event level, and within person-results in different findings, but this variation in has not been studied systematically. The proposed study offers a unique opportunity to address an important substantive and methodological issue by explicitly examining how and why different methods used to examine the relationship between drinking and condom use have produced different results. In this project, we will: (1) Combine data from several general population surveys that use the same measures (2). Describe the relationship of drinking to unsafe sex at three different levels: the person level, the event level, and within person (3). Systematically investigate reasons for differences in findings from the three methods, including the presence of nondrinkers in the samples and the role of confounding individual differences (4). Examine the role of individual and situational factors-including sexual attitudes, expectancies about drinking and sex, and perception of risk-in explaining the alcohol-risky sex link at different levels.